At Honeywell, I develop and deploy end-to-end machine learning solutions, from data preprocessing and feature engineering through evaluation, deployment, and monitoring. I analyze structured and time-series data using Python, SQL, and PySpark.
For an industrial predictive maintenance platform, I developed a machine learning solution to predict equipment failure risk from historical sensor, operational, and maintenance data. I trained and compared models including Random Forest, XGBoost, and Gradient Boosting, and exposed predictions through FastAPI REST APIs.
I also developed an enterprise AI knowledge assistant that lets users query internal documents with LLMs, RAG, and semantic search. Its document pipeline processes PDFs, stores Sentence Transformers embeddings in ChromaDB, and uses Ollama to generate grounded responses.
I use MLflow for experiment tracking and model registry, and have worked with Databricks, Delta Lake, and AWS services. I hold a B.Tech in Computer Science and Engineering from Indore Institute of Science and Technology.
